3 papers
math.ST2026
A complete characterization of testable hypotheses
Martin Larsson, Johannes Ruf, Aaditya Ramdas
We revisit a fundamental question in hypothesis testing: given two sets of probability measures and , when does a nontrivial (i.e.\ strictly unbiased) te…
math.ST2026
Testing hypotheses generated by constraints
Martin Larsson, Aaditya Ramdas, Johannes Ruf
E-variables are nonnegative random variables with expected value at most one under any distribution from a given null hypothesis. Every nonasymptotically valid test can be obtained…
math.ST2025
The numeraire e-variable and reverse information projection
Martin Larsson, Aaditya Ramdas, Johannes Ruf
We consider testing a composite null hypothesis against a point alternative using e-variables, which are nonnegative random variables such that $\mat…